We plan to build on an existing database of educational, health, and nutritional outcomes for school children in western Kenya (collected from 1998-2002), and extend it for an additional six years, in order to estimate the impact of improved child health and nutrition on long-run life outcomes. The proposed study will examine the long-run impact of a recent school-based health project - free medical treatment for intestinal helminths (worms), a serious health problem for Kenyan children - that took place in 75 rural primary schools. Medical treatment was randomly "phased in" to a subset of schools between 1998 and 2001. Schools phased into the assistance project in earlier years have served as "treatment schools", and those phased in later as "comparison schools". The proposed study will track pupils in these schools for six more years. The resulting dataset - the Kenyan Life Panel Survey (KLPS) - will contain unique longitudinal educational, labor market, health, nutritional, demographic, and cognitive information for 6,800 children over ten years (1998-2008). Our recent research indicates that deworming led to significant health, nutritional, and educational gains among pupils in the early treatment schools relative to comparison school pupils, in the first five years of the program. Deworming led to dramatic gains in school attendance and enrollment - the total number of school days missed in 1998-1999 fell by one-quarter in treatment schools - and by 2003, educational attainment was significantly higher among treatment school children. The proposed study will allow us to examine whether these educational gains persist through time, and if they translate into labor market or fertility impacts as the children enter adulthood. If there are strong links between child health gains (from deworming) and adult human capital formation and poverty, the results of the proposed study may justify increased investment in child health and nutrition programs. Large-scale, interdisciplinary, multi-use longitudinal (panel) household datasets running over ten years are very rare in less developed countries, and particularly in Africa, and there are many possible research uses for the dataset that do not rely on the deworming experiment. For instance, the KLPS panel dataset will be well-suited to examine the effect of a parent death on the life outcomes of surviving orphan children. The issue of orphan welfare is especially relevant in Kenya - and many parts of Sub-Saharan Africa - where HIV/AIDS deaths have led to a sharp increase in orphans. We plan to make the complete KLPS dataset publicly available at no cost to other investigators by the end of the project. [unreadable] [unreadable]